| 학회 |
한국고분자학회 |
| 학술대회 |
2021년 봄 (04/07 ~ 04/09, 대전컨벤션센터) |
| 권호 |
46권 1호 |
| 발표분야 |
분자전자 부문위원회 Ⅰ |
| 제목 |
Biologically Plausible Artificial Synaptic Array: Replicating Ebbinghaus’ Memory Curve with Selective Attention |
| 초록 |
The nature of repetitive learning and oblivion of memory enables humans to manage vast amounts of memory by prioritizing information for long-term storage. This paper presents an artificial synaptic array, which mimics the biological memorization process by replicating Ebbinghaus’ forgetting curve. To construct the artificial synaptic array, access transistors and synaptic memory transistors were designed using indium-gallium-zinc-oxide and poly(3-hexylthiophene), respectively. To secure the desired performance of the access transistor in regulating the input signal to the synaptic transistor, the content of gallium in the access transistor was optimized. We realized repetitive learning characterizing Ebbinghaus’ oblivion curves using an artificial synaptic array with optimized conditions for both transistor components. Furthermore, we were able to mimic selective attention for information prioritization in the human brain. |
| 저자 |
노동규1, 김성찬2, 최윤영2, 우휘제2, 강문성3, 송영재2, 안종현1, 이윤명2, 조정호1
|
| 소속 |
1연세대, 2성균관대, 3서강대 |
| 키워드 |
bio-inspired; repetitive learning; memorization; artificial synapse; multi-state; synapse array
|
| E-Mail |
|
